A major approach to the ethics of artificial intelligence is to use social choice, in which the AI is designed to act according to the aggregate views of society. This is found in the AI ethics of “coherent extrapolated volition” and “bottom–up ethics”. This paper shows that the normative basis of AI social choice ethics is weak due to the fact that there is no one single aggregate ethical view of society. Instead, the design of social choice AI faces three (...) sets of decisions: standing, concerning whose ethics views are included; measurement, concerning how their views are identified; and aggregation, concerning how individual views are combined to a single view that will guide AI behavior. These decisions must be made up front in the initial AI design—designers cannot “let the AI figure it out”. Each set of decisions poses difficult ethical dilemmas with major consequences for AI behavior, with some decision options yielding pathological or even catastrophic results. Furthermore, non-social choice ethics face similar issues, such as whether to count future generations or the AI itself. These issues can be more important than the question of whether or not to use social choice ethics. Attention should focus on these issues, not on social choice. (shrink)

Parfit's requirements for an ideal Theory X cannot be fully met since the Mere Addition Principle and Non-Antiegalitarianism imply the Repugnant Conclusion: Theory X does not exist. However, since the Repugnant Conclusion is really compelling, the Impersonal Total Principle should be adopted for impartial comparisons concerning future generations. Nevertheless, where our own interests are affected, we may yet choose to be partial, trading off our concern for future goodness with our self-interests. Theory X' meets all Parfit's requirements except the Mere (...) Addition Principle in less compelling cases. (shrink)

According to outcome welfarism, roughly, the value of an outcome is fundamentally a matterof the individual welfare it contains. I assess various suggestions as to how to spell out this idea more fully on the basis of some basic intuitions about the content and implications of welfarism. I point out that what are in fact different suggestions are often conflated and argue that none fully captures the basic intuitions. I then suggest that what this means is that different doctrines of (...) welfarism may be appropriate in different contexts and that when deciding on a particular doctrine, we need to consider which intuitions it does accommodate. Finally, I consider the issue of just how a benefit must be related to an outcome in order to contribute to its value. (shrink)

In Reasons and Persons, Derek Parfit cannot find a theory of well-being that solves the Non-Identity Problem, the Repugnant Conclusion, the Absurd Conclusion, and all forms of the Mere Addition Paradox. I describe a “Quasi-Maximizing” theory that solves them. This theory includes (i) the denial that being better than is transitive and (ii) the “Conflation Principle,” according to which alternative B is hedonically better than alternative C if it would be better for someone to have all the B-experiences. (i) entails (...) that Quasi-Maximization is not a maximizing theory, but (ii) ensures that its evaluations will often coincide with such theories. (shrink)

Welfare biology is the study of living things and their environment with respect to their welfare. Despite difficulties of ascertaining and measuring welfare and relevancy to normative issues, welfare biology is a positive science. Evolutionary economics and population dynamics are used to help answer basic questions in welfare biology : Which species are affective sentients capable of welfare? Do they enjoy positive or negative welfare? Can their welfare be dramatically increased? Under plausible axioms, all conscious species are plastic and all (...) plastic species are conscious. More complex niches favour the evolution of more rational species. Evolutionary economics also supports the common-sense view that individual sentients failing to survive to mate suffer negative welfare. A kind of God-made fairness between species is also unexpectedly found. The contrast between growth maximization, average welfare, and total welfare maximization is discussed. It is shown that welfare could be increased without even sacrificing numbers. Since the long-term reduction in animal suffering depends on scientific advances, strict restrictions on animal experimentation may be counter-productive to animal welfare. (shrink)